Visual analysis
Omni Insights Visual Analysis is the core feature module of the Omni Insights data analysis tool. It helps game operations / R&D staff analyze community data from macro (overall), micro (details), and meso (trending topics, keywords) perspectives, so they can quickly understand what players are saying.
Entry point and home page overview
Where to find it
From the top bar of the Omni Insights Platform, go to Analytics to open the Visual Analysis page.
Home page elements
-
Create Analysis: Click the button on the left to create a new analysis report.
-
Template area: The current version provides static template placeholders (preset templates will be supported in later versions).
-
Saved reports list: Shows all saved reports, including name, remarks, and last updated time, and supports the following operations:
- Click a report name or the Edit button to open the details page.
- Click the Delete button to remove a report.
- Use the search box at the top to search by report name or remarks.
Core operations
Set the data range
The data range is the basic filter of an analysis, so set it first:
Select channels and data types
-
Select the date range of the data
-
Channel selection:
- You can select multiple channels connected to the current game.
- You can select All Channels (no channel restriction).
-
Data type selection:
-
Specific types: The current version supports posts, videos, reviews, and comments.
-
Shortcuts:
- Topics: Automatically selects posts + videos + reviews.
- All: Selects all supported types.
-
Note: After you select a shortcut, you can't select specific types.
-
Configure filter conditions
Filter conditions help you pinpoint the target data. Follow these steps:
-
Select the content type:
- You can select specific types (such as posts or comments) or shortcuts (All, Topics).
- The Topics shortcut is available only when the data type includes posts / videos / reviews.
-
Select fields: Based on the content type, select the fields to filter by (such as title, sentiment, and keywords).
-
Select filter conditions: Different field types support different conditions. Common types are as follows:
- Text fields (title, content, etc.): is, is not, contains, does not contain, is not empty, is empty, RegEx match.
- Numeric fields (importance, rating, etc.): is, is not, less than, less than or equal to, greater than, greater than or equal to, range, is not empty, is empty.
- Sentiment fields: Filter by type (positive / neutral / negative) or by sentiment score range (-1~1).
- Nested fields (such as the parent topic of a comment): Select the parent field first (such as Parent Topic), then select a child field (such as Topic Title) and the condition.
Configure charts and analysis dimensions
After you set the data range and filter conditions, configure the analysis:
-
Select the chart type: The current version supports bar charts, stacked bar charts, line charts, and data tables
-
Select dimensions and sub-dimensions:
- Dimension: The field used for grouped analysis (such as channel, publish time, or sentiment type).
- Sub-dimension: Breaks the data down further (such as adding a sentiment type sub-dimension under the channel dimension).
- Note: The publish time dimension supports hour / day / week / month granularity. Numeric dimensions support discrete values or custom ranges.
-
Select metrics: The statistics used to measure the data. The current version supports:
- Content count: Counts the number of content items.
- User count: Counts the number of deduplicated user IDs.
- Advanced metrics such as scoring tags and extraction tags will be supported in later versions.
View visualization results
- The chart area uses a chart above, table below layout, with the data table showing the detailed data.
Data drill-down
- Click a data point in a chart or table to drill down and view the detailed content:
- Posts / videos / reviews: Go directly to the content details page.
Report management
Edit and delete reports
- Edit: On the report details page, change the configuration and click Save again to update the report.
- Delete: Click the Delete button in the report list on the home page, or delete the report on its details page.
Extended analysis capabilities
Visual Analysis now supports combining engagement, custom metrics, time aggregation, and richer filter conditions in the same analysis report.
Engagement and custom metrics
- Engagement metrics of posts and videos, such as views, likes, comments, shares, and favorites, can be used directly as analysis metrics, formula sub-metrics, or numeric filter conditions
- Custom metrics can be used to build business metrics such as engagement rate and content efficiency, and can also be added to custom dashboards
- Formula sub-metrics can be reordered by dragging, and you can configure up to 10 of them. Formula results are always rounded to two decimal places
Time and filters
- The publish time dimension supports aggregation by quarter and by year
- Topics, posts, and videos can be filtered by conditions such as Official and Content ID. Tag filters show tag types such as category tags, scoring tags, and extraction tags
- Dashboards can have a space-level default analysis time, ranging from the last 1–90 days, with the option to include the current day

